unloading
Master of Science Thesis in the Masters Programme Geo and Water Engineering
ARAZ ISMAIL
FITSUM TESHOME
FITSUM TESHOME
Cover: Excavation under the bridge at section E13 at Lrjeholm and horizontal
displacement readings in contour form in Plaxis. The picture was taken by Fitsum
Teshome in January 2011.
Chalmers Reproservice
Gteborg, Sweden 2011
Analysis of deformations in soft clay due to unloading
Master of Science Thesis in the Masters Programme Geo and Water Engineering
ARAZ ISMAIL
FITSUM TESHOME
Department of Civil and Environmental Engineering
Division of GeoEngineering
Geotechnical Engineering Research Group
Chalmers University of Technology
ABSTRACT
In infrastructure projects such as road and railway construction it is necessary to
perform a number of earth excavations wherever required. These excavations cause
difficulties regarding deformations and slope stability on the soil mass. The
deformation of soil is highly dependent on the stiffness of the soil. Therefore, it is of
particular interest to determine the appropriate stiffness modulus of soil in order to
study the deformation properties of soil. The objective of this thesis is to study in
depth the deformation of soft clay due to unloading or excavation.
In this project the excavation was carried out on section E13 at Lrjeholm which is
part of the undergoing new railway project between Gothenburg and Trollhttan. It
was carried out on soil layer that consisted of silty sand for the first few meters and
soft clay for deep layers below. Due to the construction conditions of the area, the
excavation was taking place under the newly constructed bridge and the existing
highway E45 near this section had to be diverted and placed under this bridge. This
thesis focuses on the deformation analysis of soft clay due to this excavation. The
deformation analysis mainly deals with the unloading modulus based on field
deformation measurements, field and laboratory tests and previously conducted
research on the unloading modulus of soft clay.
The field deformation measurements were taken from an inclinometer device, surface
point measurements and bridge point measurements installed at different location
where the excavation is taking place. The laboratory tests mainly include the
oedometer, CRS test where the two different soil stiffness moduli, M0 and ML are
obtained. However, in this case, due to unloading and the clay being slightly over
consolidated, the modulus M0 was only considered.
In this thesis, the main deformation analysis was performed through the help of the
finite element based geotechnical computer software called Plaxis. For the simulation,
two soil models, namely; Mohr coulomb and Hardening soil, were used in order to
catch the real deformation behaviour of soft clay. The results of the analysis from
these soil models are cross checked with the field deformation measurements and the
unloading modulus is estimated through back calculation. Based on this, a conclusion
is made to verify which approach gives a reasonable and realistic unloading modulus.
Key words: Clay, excavation, deformations, unloading modulus
I
Contents
ABSTRACT I
CONTENTS II
PREFACE IV
NOTATIONS V
1. INTRODUCTION 1
1.1 Background 2
1.2 Aim 3
1.3 Material and method 3
1.4 Delimitation 3
2. THEORY 4
2.1 Properties of soil 4
2.1.1 Density 5
2.1.2 Porosity 6
2.1.3 Consistency limits 6
2.1.4 Natural water content 8
2.1.5 Shear strength (Vane, cone, CPT, shear test) 8
2.1.6 Sensitivity 9
2.1.7 Effective stress and preconsolidation pressure 9
2.1.7 Permeability 10
2.2 Deformation of soft soil 10
2.2.1 Soil stiffness 11
2.2.2 Unloading modulus 12
2.2.3 Deformation determination 13
2.3 Empirics 16
2.3.1 Shear strength 16
2.3.2 Modulus 17
2.4 Plaxis 18
2.4.1 Input 18
2.4.2 Calculation 19
2.4.3 Output 19
2.4.4 Soil models 20
4. PLAXIS ANALYSIS 41
4.1 Input phase 41
4.1.1 Mohr coulomb 42
4.1.2 Hardening soil 44
4.2 Calculation phase 46
4.3 Outputs 47
4.3.1 Mohr coulomb 48
4.3.1 Hardening soil 49
5. RESULT COMPARISON 53
6. PARAMETER ANALYSIS 55
8. RECOMMENDATIONS 58
REFERENCES 59
APPENDICES 61
Greek letters
Bulk density
s Compact density
fu Reduced undrained shear strength
Correction factor
fk Unreduced shear strength
The construction of the two bridges for the railways crosses the existing highway,
E45. Due to this, the existing highway, E45, will be diverted and placed under one of
the railway bridges, see Figure 2. This involves excavation under one of the two
bridges that is already constructed. The excavation in the area will cause considerable
deformations in soft clay soil where the bridges and their foundation rest on. These
deformations will cause a movement on the bridge foundations in direction of the
excavation, especially at one end where the bridge support system is a hinge type.
During this temporary traffic stage under one of the bridges deformation
measurements were executed. Based on these measurements and with full scale
excavation test and its evaluation the deformation properties of the soil were
developed. In this thesis, this deformation analysis of soft clay due to the excavation
is performed with the help of geotechnical software called Plaxis. The results from
field deformation measurements and Plaxis deformation analysis are also compared in
order to achieve better results regarding the deformation properties of soft clay soil.
Moreover, predictions on the deformations that are expected to occur in the next main
excavation stage near section E13 can be made based on the deformation analysis of
soft clay at section E13 at Lrjeholm.
1.1 Background
The new rail way project between Gothenburg and Trollhttan involves the
construction of two new railway bridges namely USP and NSP, at section E13 at
Lrjeholm. Due to the conditions in the construction area, the existing highway E45
must be temporarily diverted and placed under one of the two new bridges. However
this process involves excavation under the bridges. The excavation includes removal
of six meters of soil that consists of siltysand for the first two meters and soft clay for
the other four meters. As consequence of this excavation, horizontal and vertical
deformations are expected.
To evaluate the deformations due to this excavation, Norconsult will perform field
and technical deformation analysis. The field analysis is performed by installing
inclinometer and point measurement devices at the actual excavation area, where
these devices measure the deformations at the different stages of the excavation. The
technical analysis involves usage of geotechnical software to simulate the deformation
due to this excavation. This thesis is part of the technical deformation analysis where
Plaxis, a finite element based advanced geotechnical software is used for the
1.2 Aim
The aim of this project is to study in depth the deformation properties of soft clay due
to excavation, and to use this analysis as a basis to decide which soil parameter has
substantial influence on the deformation of soft clay. Special attention is given to the
stiffness Modulus that will contribute much in the deformation analysis and computer
simulation. This is because, in this case, due to unloading of the soil, the stiffness
modulus is expected to be different from the one that is obtained from the oedometer
or triaxial laboratory test. Different empirical approaches are used to get a reasonable
and accurate unloading modulus. Furthermore, measurement readings of deformation
on the site are compared with output from plaxis software simulation to see which
empirical approach gives better and realistic result.
1.4 Delimitation
Section E13 at Lrjeholm or the excavation area under the bridge, where the existing
E45 will be diverted and placed under the bridge, is the area where this thesis mainly
focuses on. Only this section will be discussed in this project, and prediction on the
other excavations will be made based upon the analysis in this section.
Other important factors having a major impact on the properties of a soil are the shape
of the particles and the distribution of the size of the grains. These factors affect the
geotechnical properties of a soil, such as strength, deformation, capillarity and
permeability properties.
Clay, which is considered as a soft soil, is very common soil in southwestern part of
Sweden. Clay has a grain size less than 0.002 mm, which is a low value when
compared to the grain size of sand, which has a size between 0.062 mm, see Figure 4.
Figure 4: Grain size of clay, silt and sand. (The COMET program, 2011)
Due to the size of the grains of clay and its poor contact with each other, clay has
lower permeability and strength and lower deformation resistance than a soil with
bigger grains like sand. The particles in clay are not in direct contact with each other
when compared to other soils consisting of bigger particles. Instead the particles in
clay are adhered together by molecular attraction forces. This means that clay has an
open structure that makes it to be very compressible soil, see Figure 5. (Sllfors,
2001)
Clay that has been exposed to vey high pressures attains a dense structure and makes
the clay to be a lower compressible soil than it was before.
When excavating in a clay layer, horizontal and vertical (heave) deformations will
occur. Because of this problem, there is a challenge to construct buildings, roads,
bridges and railways on a ground with deep layers of clay. Installation of piles and
sheet piles and slope stability measures are performed as geotechnical solutions in
order to obtain stable excavations and firm foundation, when building bridges,
buildings, railways and roads in an area consisting of deep layers of soft clay.
(Sllfors, 2001)
2.1.1 Density
Density of soils, which represents the physical conditions of soil, is an important
parameter that has to be determined, since the value is different for different soil
types. Density of soils refers to the mass of small particles that constitutes the soil
mass divided by the total volume they occupy. In geotechnical laboratory test, two
types of densities are determined; bulk density and compact density, see Table 1 for
soil densities of different soil types.
The first one refers to the density of the soil measured including the pores that are
contained in the undisturbed soil sample. It is explained in equation (1) below.
( )
( )
2.1.2 Porosity
A soil consists of particles and pores, which are filled with water or/and gas as stated
in the introduction of this chapter. Porosity of soils, n is determined by dividing the
volume of the pores in a soil sample with the total volume of the soil sample, see
equation (3) below.
( )
Where, n  Porosity
Vp  Pore volume
Vtot  Total sample volume
The relation between the volume of the pores and the volume of the particles (solid
part of the soil) can be seen in equation (4).
( )
Where, e  Pore ratio
Vp  Pore volume
Vs  Particle (substance) volume
The upper two equations can be combined and equation (5) shows the relation.
( )
The liquid limit, WL and the plastic limit, Wp indicate the water content where a
cohesive soil sample has a nonplastic, plastic or liquid consistency.
The fall cone test is a Geotechnical laboratory test that is used worldwide, to
determine the consistency limits or Atterberg`s limits of soils. Among the consistency
limits, liquid limit is the parameter determined from this test. Liquid limit is the
measure of the water content of clay, at which the soil starts to change from plastic to
liquid state. It is expressed in percent, with values in range between 15 and 100%,
where this value becomes high up to 90% for highly plastic clay. Even if, liquid limit
is one parameter that is determined from the fall cone test, undrained shear strength of
the soil at the measured liquid limit is the significant parameter to be evaluated, which
then will be used together with the undrained shear strength obtained from field
investigation, to determine the actual undrained shear strength of the soil. (Houlsby,
1982)
For this test a piston soil sample is taken in the field, where these samples are taken to
the laboratory using ordinary laboratory tubes. The test is performed on undisturbed
sample for every meter until certain depth, usually up to 10m, then every two or three
meters for the rest of the investigation depth. The cone used in this test has a weight
of 10, 60, 100 or 400 gram, with an angle of 30 or 60 degrees at the tip, and it is
dropped under controlled conditions on the placed clay sample that has a smooth
contact surface. Then the tip of cone starts to penetrate the clay sample, and this
procedure will be performed three times. After each test, part of the clay sample
which is equal to1.5 times of the penetration depth is removed and put in a bowl after
each step. Finally the average of the three values is evaluated and will be considered
as the undrained shear strength of the clay. Another test, but with the same procedure,
is performed on a remoulded clay sample, which was the removed clay and put in a
bowl during the earlier test. After evaluating the shear strength from this test as well,
sensitivity analysis can be performed by comparing the shear strength from the two
tests, which will be more explained later under the sensitivity analysis. The device
used to perform this test is shown in Figure 7. (Sllfors, 2001)
A correction factor , which is a function of liquid limit, see equation (6), will be
applied in order to reduce the undrained shear strength obtained from the fall cone
test, since the values obtained from the test has shown to be higher than the actual
undrained shear strength of the clay.
( )
Where, fu  Corrected undrained shear strength, same as cuk
 Correction factor = (0.43/WL)0.45
fk  Uncorrected shear strength from fall cone test
( )
2.1.6 Sensitivity
Sensitivity test is performed in soft plastic soils that are expected to show a significant
reduction of shear strength when subjected to shearing action. As stated earlier,
sensitivity test is carried out during fall cone test, and it is defined by the ratio of the
undisturbed shear strength divided by the remoulded shear strength, see equation (8)
below. (Sllfors, 2001)
( )
Where, St  Sensitivity
fu  Undisturbed shear strength
rem  Remoulded shear strength
The relation between the effective stress and the preconsolidation is expressed as
overconsolidation ratio (OCR), which is the ratio between the preconsolidation and
the effective stress, and also as preover burden pressure (POP), which is the
difference between them. For normally consolidated soils the OCR value is 1 and the
POP is 0. These values increase for slightly and heavily over consolidated soils
accordingly.
2.1.7 Permeability
Permeability (k) or sometimes called hydraulic conductivity is the measure of the
ability of water to flow through a soil media. It has a measurement unit of m/s. Like
other soil parameters permeability of soil is dependent on the soil origin, particle size
and distribution and the bond between particles. For instance, the permeability of
gravel is higher than sand, and the permeability of sand is higher than silt and clay.
Among the listed soil types clay has low permeability due to the above mentioned
reason. Usually permeability of soil is determined from laboratory tests known as
constant head or falling head permeability test, or sometimes from field test called
percolation test. CRStests are the most common test in Sweden to determine
permeability for clay.
Figure 9: Different material models of soil behaviour: 1) linear elastic model, 2) elasticperfectly plastic
model, 3) nonlinear model, 4) elastoplastic model. (Persson, 2004)
In analysis of deformation of soft soil, the elastic deformation limit is the pre
consolidation or similar yield stress for drained conditions, and shear strength for
undrained conditions. Failure criterion is based up on Mohr coulomb failure
criterion, which is determined by performing triaxial and oedometer laboratory tests.
Figure 11: Different types of modulus, Muunloading modulus, Mccyclic modulus, Mssecant modulus,
Mrreloading modulus and Mttangent modulus. (Briaud, 2010)
By increasing the load on the sample an isotropic increasing stress can be obtained.
The test is performed in two stages, which are consolidation and shearing. The
following stressstrain relation can be obtained from the triaxial test, see Figure 13.
Figure 13: A standard drained triaxial test showing the hyperbolic stressstrain relation. (Brinkgreve et
al, 2010)
As the figure shows unloading/reloading modulus, Eur and secant modulus, E50 can be
determined, which are important deformation parameters to be used as an input values
in Plaxis deformation analysis.
2.2.3.3 Inclinometer
Horizontal deformations occur due to excavations, landslides, unstable slopes etc. A
device called inclinometer is usually used to measure horizontal deformations due to
these geotechnical problems. The instrument is installed before the construction
begins, so that initial readings are collected. The device consists of a casing, a probe, a
The casing, see Figure 15, is installed in a vertical borehole, which is situated in the
soil layer where the deformations are expected. It is installed down to a depth where
no deformations are expected to occur, and this is done so the casing will not rotate. It
can also be installed in a way that it reaches the firm bottom (bedrock). (Brouwer,
2007)
When the ground moves, the casing anchored in the borehole will also move and by
measuring the inclination and comparing to initial readings of the movement, the
deformation is obtained. Usually the measurements are done twice by pulling up the
inclinometer probe from bottom of the casing to top. The first measurement is done
parallel to the inclinometer`s wheel, which is the direction of the expected movement
of the ground (A) and the second one is performed perpendicular to the wheel (B), see
Figure 16. The movement of the ground is obtained by calculating the resultant of A
and B axis. (Brouwer, 2007)
2.3 Empirics
The different soil parameter values obtained both from field investigations and
laboratory tests should be controlled. This is due to the fact that the parameter values
from the tests have shown to be different from the one in real case. This value
difference may arise due to change in conditions of the area due to different
geotechnical works, or due to human and machine errors that occur while performing
the field and laboratory tests. For this project, deformation due to excavation on clay,
two important parameters, shear strength and stiffness modulus of clay are controlled
with reasonable empirical relations.
( )
( ) ( )
{
Where the values, A = 1500 and as is in the range of 0.0025 to 0.005.
The last approach to determine the unloading secant modulus is using the outcome
from FEanalysis performed with the MITS1 soil model. The parameter set has been
studied in detail by Kullingsj (2007). The MIT formulation is based on field tests and
laboratory tests, which shows how the unloading modulus varies with change of OCR
and vertical over consolidation pressure. It is presented in the form of chart, and it
also consists of different approaches proposed by different researchers. It is shown in
Figure 17 below. This approach and the second approach see equation (12), give
similar unloading modulus as the one stated in the tender document, see equation (11).
Due to this the unloading modulus from the tender document and the second approach
are used in Plaxis simulation, since the deformations results for all three moduli are
expected to be similar.
2.4 Plaxis
In different geotechnical applications, Plaxis, a two dimensional finite element based
computer program, is used for simulation purpose. This sophisticated computer
program is mainly used for analysis of deformation and stability. It has advantageous
feature that enables users to choose different soil models, which is dependent on
mechanical deformation behaviours of soil, for the simulation. In this project, among
the different soil models, Mohr coulomb (MC) and hardening soil (HS) models are
chosen for the deformation analysis. The basic principles and assumptions in these
models are discussed in detail later.
2.4.1 Input
The purpose of the input program is to create the geotechnical finite element model, to
define and assign material properties and to generate the initial stress with in the soil
layers and boundary condition. To create the geometry of the model, points, lines, and
clusters (areas) are used. Different types of geotechnical and structural elements can
be inserted into the model. The fixity and support system of the elements are also
defined depending on the conditions of the application. Additionally, different types
of loads and boundary conditions can be defined. The material properties definition
stage involve inserting different soil parameters like youngs modulus (E),
cohesion(c), poisons ratio (), friction angle (), and dilatancy angle () into the
2.4.2 Calculation
The calculation program is the part of the whole simulation where the analysis of the
created model is performed. The procedure is started by defining the initial stage or
the already created finite element model in the input program. Then the other
construction stages can be defined, otherwise the program only simulates the model
defined from the input program. Plaxis offers three types of calculation modes for the
users. These include, plastic, where the deformation is expected to be elasticplastic,
consolidation, where the development of or dissipation of excess pore pressure in the
clay is considered, and phic reduction, a safety analysis by reducing the strength
parameters of the soil. In this calculation part there are also load stepping and control,
and load incremental multipliers option where the user can insert appropriate value
depending on the conditions of the application. Additionally, sensitivity analysis can
be performed to determine the influence of individual parameters on the output of the
simulation. Before the final calculation is started, specific points can be selected to
generate loaddisplacement curves and stress path or stress strain curves for those
points. (Bringreve et al, 2010)
2.4.3 Output
In Plaxis, the results from the overall simulation can be obtained from the output
program. It mainly emphasizes on results of generated input data and finite element
based calculation. The outputs can be in the form of plots of full model, or plots of
special objects of the model and in the form of tables. For a specific output, the result
can be presented in the form of arrows, contours and shadings. Different types of
outputs can be obtained from the simulation, choosing the important results will
depend on the purpose of the simulation and the choice professional performing the
simulation. The outputs obtained from Plaxis include, deformed mesh of the whole
model, different types of deformations and strains and effective and total stresses. For
structural and geotechnical elements, the possible outputs from the simulation consist
of deformations, axial shear forces and bending moments. For a particular section,
where results from that section are relevant, Plaxis offers a way to view the outputs in
that section inform of crosssection. Loaddisplacement curves and timedisplacement
curves can also be generated for those specific points which were selected before
running the calculation. All the relevant outputs from the simulation can be
documented in report form. (Bringreve et al, 2010)
Figure 18: Mohr coulomb failure criterion with effective and undrained strength parameters.
(Bringreve et al, 2010)
Figure 19: The figure shows the main idea of an elastic perfectly plastic soil model. (Bringreve et al,
2010)
The calculations in a Plaxis simulation by using Mohr coulomb model are fast and
therefore the model is usually used to obtain a first view of the problem. Five different
parameters are required to perform a simulation in Plaxis software by using Mohr
coulomb. The parameters required are the following; Young`s modulus E, poison`s
ratio , friction angle , cohesion c and angle of dilantancy . The selected modulus
of elasticity, E is an average value of the modulus, which increases linearly with depth
by an increment Eincrement. The modulus of elasticity (Young`s modulus) is obtained
from a triaxial test or a CRStest. If the modulus is obtained by use of a CRStest, the
modulus, Eeod has to be transformed to an elasticity modulus, E by the equation (17).
( )
( )
( ) ( )
If effective values of elasticity modulus, E and poison`s ratio, are used in Plaxis
software to perform a simulation, the equation (18) has to be used to transform the
elasticity modulus into an effective one. (Bringreve et al, 2010)
( )
( )
The insertion of the preconsolidation pressure into the model is based on the fact that
for stresses higher than the preconsolidation pressure the stress strain curve gives a
modulus, ML that is smaller than the elastic modulus M0, see Figure 21.
Figure 21: Stiffness modulus at elastic (M0) and plastic (ML) region obtained from CRS test.
The modulus in the hardening soil model is described more accurately by adding three
moduli; Eur, which refers to reloading/unloading modulus and E50 and Eoed. E50 refers
to secant modulus from triaxial test and Eoed refers to tangent modulus that
corresponds to a modulus for stresses higher than the preconsolidation pressure
evaluated from the CRS test, see figure above. (Bringreve et al, 2010)
These moduli are stress dependent and in the following formulas the change of the
three moduli can be observed by the change of soil stress (by the depth), see the
equations (19), (20), & (21) below. (Bringreve et al, 2010)
( ) ( )
( ) ( )
In the above mentioned formulas the Eref is the reference Young`s modulus and pref is
reference stress for the stiffness and is set to 100 kPa, as a default value in Plaxis.
Usually Eurref is set to three times E50ref and Eeodref is set to be equal to E50ref as a
default value of the Plaxis software. However Plaxis cannot handle to big ratios
between E50ref and Eeodref and therefore Plaxis software will sometimes recommend a
value of one of the modulus to use. The exponent, m, is a factor that regulates the
stressdependence of the modulus. For soft soils the factor is set to 1 and for sands it
is set to value from 0.5 to 1. In the formulas it can also be noticed that Eoed is
dependent of the major principal stress 1, Eur and E50 are dependent on the minor
principal stress 3. The following formula is used to determine the horizontal stress,
see equation (22). (Bringreve et al, 2010)
( )
Figure 22 shows the vertical and horizontal stresses varying with depth and also
which depth Pref = 100 kPa corresponds to.
Poisson`s ratio for unloading/reloading and the soil pressure coefficient, K0 are other
parameters needed to describe the deformation properties of a soil by use of hardening
soil model.
Figure 23: The red circles shows the selected boreholes, 71003, FB41, 71007 and Syd.
3.2 Topography
The area, where the excavation is taking place is located on the east side of Gta
River. The area is characterised by its lowlying landscape. The level of the ground in
the specific stretch varies from +1 meter to +3 meters above sea level.
3.5.1 Density
The density is obtained by evaluating the laboratory data from two boreholes within
the actual stretch. The density is between 1.51.7 t/m3 through the entire clay layer,
see Figure 25.
5
10
Depth below the surface (m)
15
20
W(%),(710
03)
W(%),(710
25 07)
30
Water content (%)
5
10
Depth below the surface (m)
15
20
WL
(%),71003
WL
(%),FB41
25
WL
(%),71007
30
Liquid limit (%)
3.5.4 Sensitivity
The sensitivity of the clay has a value in range of 15 to 30. This shows that the clay is
sensitive, but not that much sensitive when compared with quick clay where the
sensitivity is over 50, see Figure 28.
5
10
Depth below the surface (m)
15
20
St,(71003)
25
St,(71007)
30
Sensitivity
The final trend line of all the shear strength evaluated from the different boreholes
shows, the clay layer has approximated shear strength value of 15 kPa in the
beginning which is constant for more or less 7m below and increases afterwards with
depth. However, to follow the increase of stiffness in the soil and to make the Plaxis
simulation less complicated, the clay layer is divided into two layers with shear
strength profile that corresponds with the above figure. The generated equations for
the shear strength, which is represented by the black line in the shear strength figure
in the two clay layers are as follows; Cu1 = 13 + 1/m kPa Cu2 = 27 + 1/m kPa
0
0 50 100 150 200 250 300
5
`c (71003)
10 `c (FB41)
Depth below the surface(m)
`c (71007)
`c,mean
15
`0
20
25
30
Pressure(KPa)
Figure 30: The diagram shows the insitu effective stresses and preconsilodation pressures evaluated
from CRStest.
3.5.7 Permeability
The permeability is evaluated from three boreholes within the actual stretch. The
permeability of the clay layer is between 0.5*10^  92*10^9 m/s, see Figure 31.
( )
The excavation in section E13 was performed to a depth of 6 meters below the ground
surface with a width of 30 meters. The excavation in this project is done in two stages.
The first one is 2 meters below the surface and the second is 4 meters below the first
excavation stage, see Figure 33.
The excavation is performed in a way that the slopes will have a safety factor around
1.4, which means that no ground failure will occur, see Figure 34.
A measurement program consisting of points is also installed on the USP and the NSP
with two rows of data points. The measurement points are shown in Figure 36.
An inclinometer device is installed down to the level 25 next to USP1 to measure the
horizontal movements next to the eastern bridge foundation, see Figure 37. The
measurement is performed manually in perpendicular directions, one direction along
the slope crest and one towards the excavation. The inclinometer measurement is
performed every other day during excavation and the deformation is measured for
each half meter below the ground surface, where the inclinometer is installed.
5
10
Depth below the ground surface (m)
15
Inclinometer
20
25
30
Cumulative horizontal deformation (mm)
The above presented diagram shows that there is no horizontal movement below a
depth of 27 meters down. As it can be seen the maximum horizontal deformation is
occurring around a depth of 9 meters below the ground surface. The horizontal
movement curve is showing that the deformation is decreasing from the depth of 9
meters to the upper level of the clay. This decrease in value is due to the situation
around the bridge foundation, which is discussed more in Outputs chapter.
0,000
20101118
20101128
20101208
20101218
20101228
20110107
20110117
20110127
20110206
20110216
Date
0,005
0,010
Horizontal deformation [m]
0,015
0,020
0,025
Pkt 9
0,030 Pkt 7
Pkt 8
0,035 Poly.
(Pkt 9)
Poly.
0,040 (Pkt 7)
Poly.
(Pkt 8)
0,045
Figure 39: Horizontal deformation readings with gauges for different dates.
As shown in the figure above, the geometry model consists of a section from the
existing E45, which was unloaded 50 years ago, current excavation area and four
different soil layers at different level modelled at a standard fixity and boundary
conditions. The different soil layers have different parameter values that have to be
inserted into the model at this phase. But, the type and value of soil parameters used
in this project in the two soil models is discussed independently in the subheadings
below.
Once the geometry model is clearly defined and the different soil layers are assigned
to their respective level, the next step is to create a triangular finite element mesh for
the finite element method of Plaxis simulation. The type of mesh used in the analysis
includes a medium mesh and a fine mesh. The latter mesh is used for the part where
deformation measurement is of great importance, whereas the medium mesh is used
for the rest of the areas that constitute the geometry model, see Figure 41.
Figure 41 shows the two types mesh that are used in this deformation analysis. The
fine mesh is used in the area where the inclinometer device is in position in the actual
excavation area, so that the results from the Plaxis analysis can be comparable with
the inclinometer readings. This based on the fact that the finer the mesh is the better
the accuracy of the calculation will be. For the rest of the area a medium mesh is used
in order not to make the analysis time taking.
Even though the analysis is based on undrained effective stress analysis, only youngs
modulus and poissons ratio are used as effective parameters. This is due to not being
able to get effective parameters for the other strength parameters without performing
laboratory tests. Using effective values of shear strength for higher effective stresses
will lead to a high shear strength compared to the undrained shear strength obtained
by field and laboratory tests. The use of effective shear strength for lower effective
stress states will lead to a low shear strength compared to the values of undrained
shear strength obtained by field and laboratory tests see Figure 42.
However, Plaxis offers the possibility to use the original value of the strength
parameters when these laboratory tests are not performed. Therefore, in this model the
effective shear strength (cohesion) C, is set to be equal to the undrained shear strength
Cu, and the effective friction angle , is set to be equal to the friction angle u. This
type of undrained analysis is called undrained B in Plaxis software. The unloading
modulus, which is the main focus in this thesis, is inserted into the model in the form
of oedometer modulus and four other approaches, where later on parameter analysis is
performed to determine which approach will give a reasonable unloading modulus.
For this soil model in parameter analysis chapter a comparison between analyses by
use of effective and undrained shear strength and friction angle will be performed.
Compared to the analysis in Mohr coulomb, the undrained analysis in hardening soil
is not only performed by effective youngs modulus and poissons ratio but also
effective shear strength and friction angle. This type of analysis is called undrained A
in Plaxis software. This type of undrained analysis is performed because a friction
angle equal to zero will lead to non stress dependent modulus which is not desirable.
This type of analysis will also give a modulus which is constant with depth.
Therefore, in this model the effective shear strength (cohesion) c, is set to be equal to
the effective shear strength c, and the effective friction angle , is set to be equal to
20. A higher value than 20 for friction angle will give a high value of the shear
strength, see figure in section 4.1.1 and this mean that obtained results are on the
unsafe side.
The effective shear strength is proximally 10% of the undrained shear strength or 3%
of the preconsolidation pressure and is calculated according to the equation (24)
below. (Sllfors, 2001)
( )
The unloading modulus, which is the main focus in this thesis, is inserted into the
model in the form of oedometer modulus and one other approach, see equation (12).
Later on parameter analysis is performed to determine which of these two moduli and
soil model will give a reasonable unloading modulus.
Because the excavation is not an underwater excavation the ground water inside the
excavation has to be removed so it will not affect the results. The removal of the water
inside the excavation is performed by using a feature in Plaxis called cluster dry, see
Figure 43.
In the fifth phase, see table above, the slope stability analysis is performed by an
option called phi/c reduction. The idea of this calculation type is to reduce the c and
until the soil collapses and then the values of cohesion and friction angle are recorded.
These values give at which values the soil fails and are compared to the actual shear
strength and friction angle. By the comparison a safety factor F for the slope failure is
obtained, see equation (25) below.
( )
After the calculation step, the deformations and the stability of the slope can be
evaluated in the outputs and are presented in the next chapter together with the results
from field measurements.
4.3 Outputs
Among the different output results that can be obtained from Plaxis, such as
deformations of geotechnical and structural elements, slope stability, steady pore
water pressure and development of excessive pore pressures and consolidation
parameters, this thesis is concerned with horizontal displacements due to excavation.
Additionally, since the excavation has taken place in short period time plastic analysis
is only performed rather than consolidation analysis and all results regarding
consolidation process obtained from the Plaxis simulation are therefore considered to
be not relevant to this project.
The horizontal phase displacement readings are taken in form of cross section in
points where the inclinometer and point measurement devices are installed in the
actual excavation area, see Figure 44.
Readings in form of cross section are taken in order to get the relevant displacements
due to the excavation and also to compare these displacements with the displacement
readings from the inclinometer and point measurements. This method of comparison
also helps to study the deformation properties of soft clay and is used as a method to
determine which soil parameter has significant influence on deformation of soft clay.
5,00
10,00
15,00
Depth below the surface (m)
20,00
M0,mean (CRS)
25,00
Tender
3*M0,mean (CRS)
30,00 Persson
5*M0,mean (CRS)
35,00 Inclinometer
40,00
45,00
Horizontal deformation (mm)
Figure 45: Horizontal deformation by using different unloading modulus in Mohr coulomb soil model
plotted together with inclinometer deformation reading.
As the figure shows horizontal displacement values resulted from stiffness modulus
from oedometer (CRS) test and the one that is used in the real project are considerably
higher than the horizontal displacements from the inclinometer. The horizontal
displacements by using stiffness modulus 35 times of the oedometer (CRS) and the
one from Perssons empirical formula have values that are closer to the inclinometer
deformation readings.
5,00
10,00
15,00
Depth below the surface (m)
20,00 3*250*Cuk
Inclinometer
25,00 5*Mo,mean
(CRS)
30,00
35,00
40,00
45,00
Horizontal deformation (mm)
Figure 46: Horizontal deformation by using different unloading modulus in Hardening soil model
plotted together with inclinometer deformation reading.
The above figure shows horizontal displacement readings from the approach that uses
the unloading modulus as 5 times of M0 from the oedometer (CRS) test. The second
approach uses the unloading modulus as 3 times of 250 times the undrained shear
strength. For the first 15 meters of the soil layer, this approach gives lower horizontal
displacements when compared to inclinometer deformation readings.
The modelled geometry of the excavation under the bridge done in Plaxis so far does
not include the foundation of the bridge. Therefore, an extra model that consists of a
diaphragm wall (DWall) installed at the inclinometer position to see how the
installation of this diaphragm wall will affect the horizontal deformation that occurs
due to the excavation, see Figure 47. The Diaphragm wall is installed up to a depth of
Figure 47: Horizontal displacement reading when the diaphragm wall is installed at the position of the
inclinometer and the installed diaphragm wall is shown highlighted in green.
The above presented diagram shows that there is a difference in value and shape of
the horizontal displacement curves between the inclinometer and the modeled cases in
Plaxis. The decreasing rate of the deformation curves is higher for the first 7m for the
inclinometer readings than for the simulated models. This is because the foundation
bridge acts like a passive force and pushes the soil back. If the foundation of bridge
was not in place in the analyzed section, the deformation curve from the inclinometer
would probably be similar to the simulated models in Plaxis, see Figure 48.
5,00
10,00
15,00
Depth below the surface (m)
Inclinometer
25,00 Hardening
(5*M0,mean,CRS)
30,00
35,00
40,00
45,00
Horizontal deformation (mm)
Figure 48: The deformation curve for the inclinometer without bridge foundation in the studied section
compared to the modeled sections for hardening soil model.

0 50 100 150 200 250 300
5,00
10,00
15,00
Depth below the surface (m)
20,00
M0,mean (CRS)
25,00 Tender
3*M0,mean (CRS)
30,00 Persson
5*M0,mean (CRS)
Inclinometer
35,00
3*250*Cuk
5*Mo,mean (CRS)
40,00
45,00
Horizontal deformation (mm)
Figure 49: Horizontal deformation by using different unloading modulus in Mohr coulomb soil model
(the five first names in the diagram) and Hardening soil model (the last two names in the diagram)
plotted together with inclinometer deformation reading
From Table 6 it can be seen that the simulation in Mohr coulomb model by use of 5
times M0 as unloading modulus gives almost the same horizontal deformations as the
measured one by the inclinometer. It can also be seen that in hardening soil model the
use of unloading modulus as 5 times M0 from CRS test gives deformations that are
similar to the inclinometer readings.
In Table 7 the horizontal deformations from the point measurements in point 9 are
compared to the surface deformations obtained in Plaxis by using different
approaches. This point is selected because of its closer distance to the simulated
section in Plaxis.
Table 7: Deformation comparison between point measurements and outputs from Plaxis
The results obtained from Hardening soil model show that using 3 times 250 times the
undrained shear strength as the unloading modulus leads to horizontal displacements
that are lower than what they are in reality. Therefore a factor of two is better to use
Brouwer, JJ.M. (2007): Insitu soil testing. Delft University of Technology, {The
Netherlands}, {2007}, 144 pp. Available at : http://www.conepenetration.com
[Accessed 20110213]
Houlsby, G.T. (1982): Theoretical analysis of the fall cone test. Available at
http://wwwcivil.eng.ox.ac.uk/people/gth/j/j3.pdf[Accessed 20110227]
Persson, J. (2004): The unloading modulus of soft clay: A field and laboratory study.
Licentiate thesis. Department of Geo Engineering, Chalmers University of
Technology, Gteborg, Sweden, 2004, 127 pp.
The comet group program (2011), University Corporation for Atmospheric Research,
Colorado, {USA}, {2011}, Available at: www.meted.ucar.edu [Accessed 201102
13]